DESCRIPTION: (applicant's abstract) The ability to monitor physiological processes in vivo with low frequency electron paramagnetic resonance (EPR) spectroscopy would contribute significantly to biomedical research. Several labs have performed preliminary experiments but have not achieved an adequate signal-to-noise ratio (SNR) for performing useful imaging. Better penetration of the body and lower energy deposition makes low radio frequencies (RF) advantageous for biomedical EPR. The low SNRs obtained by these researchers and statements made in the literature have reinforced an assumed high order dependence of SNR on frequency. Our recent calculations predict less SNR penalty at low frequency than generally assumed, thus we predict a better inherent SNR than that achieved so far. We have also developed a new kind of resonator, a crossed-loop resonator (CLR), that has shown strong advantages at microwave frequencies (ca. 3 GHz). From basic considerations, we also expect the CLR to have these advantages at low RF frequencies. We propose to experimentally test these new predictions for frequency dependency of SNR and the low frequency performance of the CLR. We will test the traditional loop-gap resonator (LGR) at frequencies from 9 GHz to 200 MHZ. We will also test the LGR and CLR under the same conditions at 200 MHZ to determine the relative benefits of each. Combinations of noise source, sample size, and rf power exist for which we predict significant SNR increase by use of the CLR. The signal and noise information obtained by the proposed work will provide a solid basis for predicting the utility of low frequency imaging for various measurement goals.